See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228625097

Lateral Response Evaluation of Single Piles Using Inclinometer Data

Article  in  Journal of Geotechnical and Geoenvironmental Engineering · December 2006

DOI: 10.1061/(ASCE)1090-0241(2006)132:12(1566)

CITATIONS READS
25 3,076

2 authors, including:

San-Shyan S. Lin
National Taiwan Ocean University
62 PUBLICATIONS   445 CITATIONS   

SEE PROFILE

All content following this page was uploaded by San-Shyan S. Lin on 04 January 2014.

The user has requested enhancement of the downloaded file.

 Lateral Response Evaluation of Single Piles
Using Inclinometer Data
San-Shyan Lin1 and Jen-Cheng Liao2

Abstract: In an effort to develop an efficient method for interpretation of lateral pile load test results via measured inclinometer data
only, an analytical model is proposed based on energy conservation of a pile-soil system. A Fourier series function is used to represent
deflection behavior of the pile-soil system. In order to obtain shear, moment, and soil reaction along the pile shaft, convergence of the
series after differentiation is guaranteed by applying the Cesaro sum technique. The concrete cracking effect is also incorporated into the
pile model to account for yielding of the pile itself. Three full-scale pile load cases are then used to verify the feasibility of the developed
methodology as well as make comparison to other methods.
DOI: 10.1061/共ASCE兲1090-0241共2006兲132:12共1566兲
CE Database subject headings: Lateral loads; Cracking; Concrete piles; Data analysis; Evaluation.

Introduction deflection versus depth profile over a range of loading. However,

Use of piles to resist lateral loading has attracted a lot of attention
from geotechnical engineers in the past decade. However, the
response of a laterally loaded pile is a complicated soil–structure
interaction problem. For concrete piles, an added difficulty is the
nonlinear behavior of concrete near structural failure. The lateral
load behavior of single piles is often obtained via in-situ lateral
pile load tests. In general, if detailed information involving p-y
curve parameters is required, many strain gauges are often in-
stalled along the length of pile shaft to develop the bending mo-
ment versus depth relationship 共Pinto et al. 1999兲. In this paper,
the use of inclinometers is recommended for lateral pile load tests
in order to obtain more detailed information for back-calculating
p-y curves. The use of slope inclinometers to measure lateral
deflection of the pile has some advantages compared to the strain
gauges, which are easily damaged during pile installation or by
improper handling. Additionally, inclinometer measurements are
less costly compared to other alternatives. However, inclinometer
data are prone to reduction errors because the slope distribution
must be differentiated three times in order to obtain p-y curves
共Brown et al. 共1994兲; Pinto et al. 1999兲.
Brown et al. 共1994兲 proposed a method for interpreting lateral
load tests. The method utilizes a least-squares regression tech-
nique that will converge to a solution for analytical p-y curves
that produce minimum error between the predicted and measured
1 sults from three full-scale pile load cases are then used to dem-
Professor, Dept. of Harbor and River Engineering, National Taiwan
Ocean Univ., Keelung 20224, Taiwan 共corresponding author兲. E-mail:
sslin@mail.edu.tw
2
Post-doctoral Student, Dept. of Harbor and River Engineering,
National Taiwan Ocean Univ., Keelung 20224, Taiwan.
Note. Discussion open until May 1, 2007. Separate discussions must
be submitted for individual papers. To extend the closing date by one
month, a written request must be filed with the ASCE Managing Editor.
The manuscript for this paper was submitted for review and possible
publication on August 24, 2004; approved on June 7, 2006. This paper is
part of the Journal of Geotechnical and Geoenvironmental Engineer-
ing, Vol. 132, No. 12, December 1, 2006. ©ASCE, ISSN 1090-0241/
2006/12-1566–1573/$25.00.
the analysis requires use of a computer code, COM624 共Wang
and Reese 1991兲, to predict deflection. In addition, the shape of
the p-y curve must be assumed and iteration over two or more
soil parameters are needed for calibration of p-y curves. A method
that needs iteration on a single parameter, the modulus of lateral
soil reaction, was proposed by Pinto et al. 共1999兲. However, this
method assumes that the initial slope of the p-y curve increases
linearly with depth, which in general can only be used for uni-
form sands.
A new method is proposed in this paper in an effort to develop
a simple yet accurate method for interpreting lateral single-pile
load test results based only on measured inclinometer data. The
pile–soil interaction response is idealized as a beam on a Winkler
spring medium. Fourier series functions are used as the deflection
function of the pile–soil system. Although application of the Fou-
rier series in structural mechanics problems such as beams or
plates is generally used for buckling or free-vibration analyses,
applications to stress analyses have also been provided by Chen
et al. 共1996兲 and Wang and Lin 共1999兲. In order to derive the
slope, shear, moment, and soil reaction of the pile, convergence of
the series after differentiation is guaranteed by applying the
Cesaro sum technique 共Hardy 1949兲. Additionally, consideration
of the nonlinear behavior of the reinforced concrete piles is in-
cluded in the proposed method. In this paper, the authors first
present the theory and development of the proposed method. Re-
onstrate application of the developed methodology and make
comparison to other methods.
Theory
In the following sections, the deflection function of laterally
loaded piles is derived based on the principal of conservation of
energy and the application of the calculus of variation. Measured
inclinometer data from lateral pile load tests, which implicitly

1566 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / DECEMBER 2006

 冕 冕
L L
1 1
⌸= EI共y ⬙兲2dz + p共z兲ydz − Hy共L兲 共4兲
2 0 2 0

Subsequently, by applying the calculus of variations to Eq. 共4兲,

the following Euler-Lagrangian governing equation for the pile–
soil system can be obtained 共Han and Frost 2000兲

EIy IV + p共z兲 = 0 共5兲

Fig. 1. Schematic diagram of soil–pile system
include pile–soil interaction effects, were utilized to determine the
coefficients of the derived deflection function. Once the deflection
function is established, bending moment, shear force, and p-y
curves at various depths along the pile can be reduced by differ-
entiating the deflection function. Assumptions of the derivation
are:
1. Soil response is idealized as one-dimensional;
Deflection Functions
Two different boundary conditions of long piles are considered as
shown in Fig. 2. The analysis presented in the following only
applies to long piles. A long pile is defined with the ratio of the
pile length versus the relative stiffness factor 共pile stiffness/
coefficient of subgrade reaction兲0.2, larger than 4 共Prakash and
Sharma 1990兲. The first case 共case I兲 considers the free head and
fixed toe condition, which represents long piles without caps
共case I兲. The other case 共case II兲 considers fixed with sway at
head and fixed at toe condition, which represents long piles con-
strained with caps at the pile head.
By applying the Rayleigh-Ritz method, the general form of the
deflection that satisfies Eq. 共5兲 can be determined by the follow-
ing Fourier series as
n
y共z兲 = 兺
i=1
Bi共1 − cos N̄␲z兲 共6兲

2. Pile material behavior is assumed as nonlinear elastic; and

3. Lateral loading is applied only at pile head.
where N̄ = 2i − 1 / 2L and N̄ = i / L represent the boundary conditions
Energy Method of case I and case II, respectively. Coefficients Bi can be deter-
mined by substituting Eq. 共6兲 into Eq. 共4兲 and minimizing the
The total energy 共⌸兲 for a pile embedded in a soil medium is function 共⳵⌸ / ⳵Bi = 0兲.
given as Alternatively expanding Eq. 共6兲, computed deflection Yc at
depth z can be presented as Eq. 共7兲
⌸=U+V 共1兲
where U⫽strain energy stored in the pile; and V⫽potential energy
of soil–pile system caused by external loads. Yc = A · B 共7兲
As shown in Fig. 1, the strain energy of the pile can be ex-
pressed as 共Han and Frost 2000兲
where

冕
L
1
U= EI共y ⬙兲 dz
2
共2兲
2 0

where E⫽elastic modulus of the pile; I⫽moment of inertia of the

pile; y⫽lateral deflection of the pile; and z⫽depth along the pile
shaft.
The potential energy of the soil–pile system caused by external
loading applied instantaneously can be expressed as 共Han and
Frost 2000兲

冕
L
1
V= p共z兲ydz − Hy共L兲 共3兲
2 0

where p共z兲⫽soil reaction at depth z; H⫽lateral forces applied at

the pile head; y共L兲⫽pile deflection at the pile head; and
L⫽embedded length of the pile.
Substituting Eq. 共2兲 and 共3兲 into Eq. 共1兲, the total potential
energy of the soil–pile system can be presented as 共Han and Frost
2000兲 Fig. 2. Boundary conditions considered in the paper

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / DECEMBER 2006 / 1567

 冤 冥
共1 − cos N̄n␲ · z1兲 . . . 共1 − cos N̄2␲ · z1兲 共1 − cos N̄1␲ · z1兲
共1 − cos N̄n␲ · z2兲 . . . 共1 − cos N̄2␲ · z2兲 共1 − cos N̄1␲ · z2兲
Aj⫻n =
· ... · ·
共1 − cos N̄n␲ · z j兲 . . . 共1 − cos N̄2␲ · z j兲 共1 − cos N̄1␲ · z j兲

and ␲ 4 1
Ip = r + 共n − 1兲Asrs2 共14兲

冦冧
Bn 4 c 2
. where rc⫽radius of the pile shaft; and rs⫽distance from centroi-
Bn⫻1 = dal axis of gross section to centroidal axis of steel rebar.
B2
Cracking of concrete will reduce the moment of inertia of the
B1 pile shaft. Applying the effective moment of inertia concept in
The least-squares rule is then applied to perform regression to ACI 共1995兲, when the moment M ⬍ M cr, the effective moment of
solve Bn⫻1 using procedures described below: inertia is
1. Obtain the square summation of the error, S, between calcu-
Ie = I p 共15兲
lated and measured deflection, Ym 关Eq. 共8兲兴
or when M cr ⬍ M ⬍ M u, the effective moment of inertia 共Ie兲 is

冉 冊 冋 冉 冊册
S = 共A · B − Ym兲T共A · B − Ym兲 共8兲
3 3
M cr M cr
2. For minimum error, Eq. 共8兲 has to be applied and used to Ie = Ig + 1 − Icr 共16兲
Ma Ma
solve Bn⫻1 for say, Ym optimized values of coefficients Bn
where M a⫽local maximum moment along the pile shaft;
⳵S M cr = f rIg / y t⫽cracking moment; M u⫽ultimate moment;
= 2AT · A · B − 2AT · Ym = 0 共9兲
⳵B f r = 2冑 f ⬘c ⫽tensile strength of the concrete; Ig⫽moment of inertia
Based on Euler’s beam theory, once the pile deflection func- of gross concrete section about central axis; y t⫽distance from
tion is determined, the moment M共z兲, the shear force V共z兲, and centroidal axis of gross section to extreme fiber in tension; and
the soil reaction p共z兲 of the pile along the shaft can easily be Icr⫽moment of inertia of cracked transformed concrete section.
obtained as given in the following equations The local pile bending moment can be estimated using curva-
ture, deduced from measured displacement from inclinometer,
d2y共z兲 and from pile flexural rigidity, EI. Depending on whether pile
M共z兲 = EI 共10兲 section is cracked or uncracked, I is the gross moment of inertia Ig
dz2
for uncracked sections; and I is the effective moment of inertia Ic
for cracked section.
dM
Y共z兲 = 共11兲
dz
Regularization with Cesaro Sum Technique

dV Since the deflection functions of the pile–soil system given in

p共z兲 = 共12兲 Eq. 共6兲 are Fourier series functions, these functions will cause
dz
ill-posed conditions after differentiation. The Cesaro sum tech-
nique is used to guarantee the convergence of these Fourier series
functions.
Nonlinear Behaviors of the Piles
The general 共C , r兲 Cesaro sum is defined as 共Hardy 1949兲

冉兺 冊
Reinforced concrete piles exhibit nonlinear behavior under load. n n+r−1 n+r−2 r r−1
Consideration of this nonlinear behavior is important to interpret Cr−1 s0 + Cr−1 s1 + . . . + Cr−1 sn−1 + Cr−1 sn
Sn = 共C,r兲 ai ⬅
the test results correctly. Based on the ACI code 共1995兲, Reese i=0 Crn+r
and Van Impe 共2001兲, and Ooi and Ramsey 共2003兲, the effective
elastic modulus of a reinforced concrete pile 共E p兲 can be ex- 共17兲
pressed as where Crn = n ! / 关r ! 共n − r兲 ! 兴, and the partial sum is si = 兺i=0
n
ai共z兲,

冋
E p = 1 + 共n − 1兲
As
Ac
册Ec 共13兲
where ai共z兲 = Bi共1 − cos N̄␲z兲.
For computational convenience, the si terms are changed to ai
terms, and can be changed to the conventional Cesaro sum as

冉兺 冊
in which n⫽Es / Ec ; Es⫽elastic modulus of the steel; Ec⫽elastic n
modulus of the concrete and is equal to 15,000冑 f ⬘c 共kg/ cm2兲 共ACI Sn = 共C,1兲 ai ⬅
s0 + s1 . . . + sn−1 + sn
共18兲
1995兲; As⫽area of the steel; Ac⫽area of concrete; and i=0 n+1
f ⬘c ⫽compressive strength of the concrete.
Based on the parallel-axis theorem, moment of inertia 共I p兲 for n n
1
the pile shaft cross sections can be expressed as 共Reese and Van
Impe 2001兲
共C,1兲 兺
i=0
ai ⬅
n + 1 i=0
兺
共n − i + 1兲ai 共19兲

1568 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / DECEMBER 2006

 Similarly, the Cesaro sum of 共C , 2兲, 共C , 3兲, and 共C , 4兲 is

Sn = 共C,2兲 冉兺 冊
i=0
n

ai ⬅
1
共n + 1兲共n + 2兲 i=0
兺
n

共n − i + 1兲共n − i + 2兲ai

共20兲

Sn = 共C,3兲 冉兺 冊 n

i=0
ai ⬅
1
共n + 1兲共n + 2兲共n + 3兲 i=0
兺
共n − i + 1兲
n

⫻共n − i + 2兲共n − i + 3兲ai 共21兲

Sn = 共C,4兲 冉兺 冊n

i=0
ai ⬅
1
共n + 1兲共n + 2兲共n + 3兲共n + 4兲 i=0
共n − i + 1兲
n

兺
⫻共n − i + 2兲共n − i + 3兲共n − i + 4兲ai 共22兲 Fig. 3. Deflection, bending moment, shear force, and soil resistance
distributions along the pile shaft of WDOT test pile
Or for general integer order r, the Cesaro sum can be ex-
pressed as

Sn = 共C,r兲 冉兺 冊n

i=0
ai ⬅ 兺
n
共n兲 ! 共n + i − n兲!
i=0 共n + i兲 ! 共n + r兲!
ai 共23兲 4. Depending on the local bending moment, update flexural ri-
gidity at each depth of the pile; and
Based on this regularization technique, the series representa- 5. Use Eqs. 共24兲 and 共28兲 to obtain the p-y curves.
tions for displacement y共z兲, slope ␪共z兲, moment m共z兲, shear force
␯共z兲, and soil reaction p共z兲 are expressed in the sense of the Ce-
saro sum as Examples

y共z兲 = 共C,1兲 冋兺n

i=1
Bi共1 − cos N̄␲z兲 册 共24兲
The use of the technique described above is illustrated using three
real examples from field load tests and compared to the results
obtained using other methods.

␪共z兲 = y ⬘共z兲 = 共C,2兲 冋兺 i=1

n

Bi · N̄ · 共sin N̄␲z兲 册 共25兲

Washington DOT Lateral Pile Load Test
A full-scale lateral load test of pipe piles was conducted by Wash-
ington DOT and was reported by Kramer 共1991兲. The 24-m-long

M共z兲 = EIy ⬙共z兲 = 共C,3兲 冋兺 i=1

n

EI · Bi · N̄2 · 共cos N̄␲z兲 册 共26兲

pile, with diameter D = 0.476 m, was embedded into deep deposit
of soft clay and was subjected to lateral load with magnitude 45,
85, 178, and 222 kN applied at head. Case I boundary condition
and constant EI through entire pile length are assumed to perform

V共Z兲 = EIy ⵮共z兲 = 共C,4兲 冋兺 n

i=1
EI · Bi · N̄3 · 共− sin N̄␲z兲 册 共27兲
analysis.
Given the boundary conditions of the pile, we can fit the de-
flection function 关Eq. 共24兲兴 to the measured inclinometer data at
the loading magnitude of 222 kN as shown in Fig. 3. Subse-

冋兺 册
n
quently, we substitute inclinometer measurement data into Eq. 共8兲
and then determine B from Eq. 共9兲. The best fit function of this
p共z兲 = EIy ⬙⬙共z兲 = 共C,5兲 EI · Bi · N̄4 · 共− cos N̄␲z兲 共28兲
i=1
case is then obtained and is expressed in the following form:

in which N̄ = 2i − 1 / 2L and N̄ = i / L are for the boundary conditions

case I and case II, respectively.
Table 1. Coefficients Bi of Washington DOT Lateral Pile Load Test
Numerical Procedures 共Loading Magnitude⫽222 kN兲
1. Apply the boundary conditions of the pile and then compute Coefficients Bi Value Coefficients Bi Value
the pile–soil deflection function from Eq. 共6兲; B1 1.98⫻ 10 −1
B6 −2.52⫻ 10−4
2. Substitute inclinometer measurement data into Eq. 共8兲 and B2 −5.30⫻ 10−2 B7 4.32⫻ 10−4
then determine B from Eq. 共9兲;
B3 1.99⫻ 10−2 B8 −5.81⫻ 10−4
3. Use the deflection function obtained from procedure 2 and
B4 −6.43⫻ 10−3 B9 5.44⫻ 10−4
utilize the Cesaro sum technique in Eqs. 共26兲–共28兲 to obtain
B5 1.43⫻ 10−3 B10 −3.61⫻ 10−4
the moment, shear, and soil reaction along pile shaft;

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / DECEMBER 2006 / 1569

 Fig. 5. Lateral load versus pile head deflection of WDOT case

Fig. 4. p-y curves of WDOT test pile ⫻

15␲
2L
冉 15
B8 sin ␲z +
2L
2⫻3
冊冉
10 ⫻ 11
⫻
17␲
2L
冊17
B9 sin ␲z
2L
冉 冊
y共z兲 =
10
冉 冊冉 1
B1 1 − cos ␲z +
9
冊冉 冊冉 3
B2 1 − cos ␲z 冊 + 冉 1⫻2
10 ⫻ 11
冊
⫻
19␲
2L
冉 19
B10 sin ␲z
2L
冊
10 2L 10 2L

+冉 冊冉8
10
5
B3 1 − cos ␲z +
2L
冊冉 冊冉
7
10
7
B4 1 − cos ␲z
2L
冊 Following similar procedures, the convergent solutions for the
third and fourth derivative of the function can be obtained simi-

冉 冊冉 冊冉 冊冉 冊
larly by selecting different Cesaro orders.
6 9 5 11 Giving the pile flexural rigidity at each depth of the pile and
+ B5 1 − cos ␲z + B6 1 − cos ␲z
10 2L 10 2L utilizing the Cesaro sum technique in Eq. 共26兲–共28兲, the moment,

冉 冊冉 冊冉 冊冉 冊
shear, and soil reaction along the pile shaft can then be deter-
4 13 3 15 mined. Fig. 3 shows the variations of deflection, bending moment,
+ B7 1 − cos ␲z + B8 1 − cos ␲z
10 2L 10 2L shear force, and soil resistance distributions along the pile shaft

+冉 冊冉2
10
17
B9 1 − cos ␲z +
2L
冊冉 冊 冉
1
10
19
B10 1 − cos ␲z
2L
冊 from either analytical results or instrumentation measurement.
From Fig. 3, the maximum bending moment was found at 5 m
below ground surface and the corresponding shear force is zero.
where the Bi 共i = 1 to 10兲 are given in Table. 1. Applying Eq. 共25兲, Fig. 4 shows the p-y curves obtained from the proposed method
the first derivative of the deflection function is given as and from Brown et al. 共1994兲 at three different depths for com-

冉 冉 冊冉 冊
冊
parison. Reasonable agreement between two results was found.
10 ⫻ 10 ␲ 1 9 ⫻ 10 Good agreement between the predicted and measured load-
y ⬘共z兲 = ⫻ B1 sin ␲z +
10 ⫻ 11 2L 2L 10 ⫻ 11 deflection relationship at the pile head is also given in Fig. 5.

⫻
3␲
2L
冉 冊冉 冊 冉 冊
3
B2 sin ␲z +
2L
8⫻9
10 ⫻ 11
⫻
5␲
2L
5
B3 sin ␲z
2L Lateral Shafts Test of the High Speed Rail Project in

冉 冊 冉 冊冉 冊
Taiwan
7⫻8 7␲ 7 6⫻7
+ ⫻ B4 sin ␲z + For the purpose of optimizing the design of pile foundations for
10 ⫻ 11 2L 2L 10 ⫻ 11
the high speed rail system, full-scale load tests were conducted on
⫻
9␲
2L
冉 冊冉 冊 冉 冊
9
B5 sin ␲z +
2L
5⫻6
10 ⫻ 11
⫻
11␲
2L
11
B6 sin ␲z
2L
two pile groups in Chiayi, Taiwan, in 1997 共Chen 1997兲. In ad-
dition to lateral load tests on the two pile groups, lateral load tests

冉 冊 冉 冊冉 冊
were also conducted on single piles B2 共bored pile兲 and P13
4⫻5 13␲ 13 3⫻4 共precast concrete pile兲, by loading the piles against the adjacent
+ ⫻ B7 sin ␲z +
10 ⫻ 11 2L 2L 10 ⫻ 11 pile caps. Only lateral pile test results of the single piles B2 and

Table 2. Soil Properties of the Test Site at Chiayi of the High–Speed Rail Project, Taiwan
Depth
Layer 共m兲 SPT-N Classification Description
1 0–3 1⬃5 ML/SM Fine sandy silt, yellowish brown, loose with some clayey silt.
2 3–8 8 ⬃ 19 SM Silty fine sand, gray, medium dense, occasionally with sandy
silt layers.
3 8–12 4 ⬃ 12 CL Silty clay, grayish brown, medium stiff, occasionally with
coarse sand seams.
4 12–16 15⬃ 29 SM Silty fine sand, gray, loose with fine sandy silt layer.
5 16–22 11⬃ 23 CL/SM Clayey silts, gray, medium dense with little sandy silts.
6 22–32 9 ⬃ 27 CL Silty clay, gray, very stiff with little fine sand.
7 32–40 14⬃ 45 SM Silty clay, gray, very stiff with silty fine sand layer.

1570 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / DECEMBER 2006

Table 3. Structure Properties of Test Piles P13 are studied in this paper. Structural properties of the tested
Bored pile PC pile piles and typical soil properties at the test site are given in Tables
Item 共B2兲 共P13兲 2 and 3, respectively.
Case I for boundary conditions of the free-head case was as-
Pile diameter 1,500 Precast 800 outside,
sumed for the single-bored pile B2 and PC pile P13. Derived
共mm兲 560 inside with
concrete infill
deflection, moment, shear, and soil resistance along the pile shaft
versus embedded depth for pile B2 are shown in Fig. 6. Numeri-
Pile length 34.9 34.0
共m兲 cal studies using the computer program LPILE 共Reese et al. 2000兲
are also conducted using the material properties provided in
Cross-sectional area 17,672 Solid: 5,027
共cm2兲 Hollow: 2,564 Tables 2 and 3. These material properties were used as input data
in the LPILE program. Reasonable agreement for the results ob-
Concrete compressive strength, f c⬘ 27.47 Precast: 78.48
共MPa兲 Infill: 20.6 tained from the proposed method and from the computer program
LPILE is found in the figure. It is worth pointing out that the
Reinforcement Yield stress, f y 471 Precast: 1,226.25
共MPa兲 Infill: 471 measured inclinometer data is the only information needed in the
proposed method. In addition, the pile deflection computed with

Fig. 6. Deflection, bending moment, shear force, and soil resistance Fig. 9. Deflection, bending moment, shear force, and soil resistance
distributions along the pile shaft of single-bored pile B2 distributions along the pile shaft of single PC pile P13

Fig. 7. Lateral load versus pile head deflection of single-bored pile Fig. 10. Lateral load versus pile head deflection of single PC pile
B2 P13

Fig. 8. p-y curves of single bored pile B2 Fig. 11. p-y curves of single PC pile P13

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / DECEMBER 2006 / 1571

Fig. 12. Arrangement of Osaka Bay test site 共adapted from Matsui
1993兲

the proposed function can also well fit the measured deflection
along the depth. The measured versus predicted pile head load-
deflection relation, with or without adjusting the flexural rigidity
of the pile in the analysis, is given in Fig. 7, in which neglecting
variation of pile flexural rigidity tends to overestimate the pile
capacity. The p-y curves obtained from the proposed method and
from LPILE also show good agreement as illustrated in Fig. 8.
Similar studies were also conducted for the PC pile P13 and the
results are shown in Figs. 9–11. Reasonable agreement for the
results of moment, shear force, and soil resistance obtained from
the proposed method and from the computer program LPILE is
also shown in Fig. 9. The importance of taking into account the
effect of pile flexural rigidity in the analysis is also shown in the
pile head load-deflection relationship of Fig. 10 and the p-y
curves in Fig. 11.
Osaka Bay, Japan, Lateral Shaft Load Test
A full-scale lateral load test of two drilled shafts was conducted at
Osaka Bay, Japan and was reported by Matsui 共1993兲. Both shafts
are 1.0 m in diameter and 25 m in length. The top 7 m of the test
site consists of alluvial sandy soil overlying an alluvial clay soil
layer of about 18 m thick. The shafts are supported on the diuvial
sandy gravel layer underlying the clay layer. The elastic modulus
Fig. 14. Deflection, bending moment, shear force, and soil resistance
distributions along the front pile shaft of Osaka Bay test pile
of the shaft is 3.0⫻ 107 kPa. The unconfined compressive
strength of the concrete is 4.0⫻ 104 kPa. Inclinometers were in-
stalled along the shaft as shown in Fig. 12.
Lateral load with magnitude of 2.1 MN was applied at a con-
nected beam, which was used to fix the heads of two shafts.
Hence, case II boundary condition is assumed for both shafts to
perform the analysis. Derived deflection, moment, shear, and soil
resistance along the pile shaft versus embedded depth for back
and front test shaft are shown in Figs. 13 and 14, respectively. As
shown in the figures, pile group effect may have caused the dif-
ference on deflections, moments, and soil resistances of both
shafts. The p-y curves obtained from the proposed method for
back and front test shafts are also given in Figs. 15 and 16, re-
spectively, for comparison.

Fig. 15. p-y curves of back pile of Osaka Bay test pile

Fig. 13. Deflection, bending moment, shear force, and soil resistance
distributions along the back pile shaft of Osaka Bay test pile Fig. 16. p-y curves of front pile of Osaka Bay test pile

1572 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / DECEMBER 2006

Summary and Conclusions
In this paper, a relatively simple method for interpreting lateral
pile test results was proposed. Background theories, as well as
detailed derivation of the proposed analytical method were de-
scribed in this paper. Moreover, the feasibility of the developed
method was further verified using results of three real case histo-
ries. Advantages of the proposed analytical method include: 共1兲
only inclinometer data is needed for deriving the deflection func-
tion; 共2兲 the derived deflection function can be used to describe
lateral behaviors of a single pile system satisfying either free or
fixed with sway pile head boundary conditions; and 共3兲 nonlinear
behavior for variation of the flexural rigidity of the pile is consid-
ered in the analysis and adjusted for each load increment.
It is necessary to point out that for short piles with high lateral
loading, permanent lateral displacement might occur at the bot-
tom of piles. Definitions of strain energy and potential energy
used in this study need to be modified in order to establish the
energy conservation relation for this case. Furthermore, energy
conservation of the soil–pile system can no longer be included
using only inclinometer data. Further study, is needed to develop
analytical model for short pile applications.
Acknowledgments
This study was supported by the National Science Council,
Taiwan, under Grant No. NSC 90-2211-E-019-023. Grateful ap-
preciation is expressed for this support.
References
American Concrete Institute 共ACI兲. 共1995兲. “ACI building code require-
ments for structural concrete and commentary.” ACI 318-95, Detroit.
Brown, D. A., Hidden, S. A., and Zhang, S. 共1994兲. “Determination of
p-y curves using inclinometer data.” Geotech. Test. J., 17共2兲,
150–158.
JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / DECEMBER 2006 / 1573
View publication stats
Chen, C. H. 共1997兲. “Data for planned pile group tests at Chiayi test site:
Workshop report.” Dept. of Civil Engineering, National Taiwan Univ.,
Taipei, Taiwan.
Chen, J. T., Hong, H. K., Yeh, C. S., and Chyuan, S. W. 共1996兲. “Integral
representations and regularizations for a divergent series solution of a
beam subjected to support motions.” Earthquake Eng. Struct. Dyn.,
25共9兲, 909–925.
Han, J., and Frost, J. D. 共2000兲. “Load-deflection response of transversely
isotropic piles under lateral loads.” Int. J. Numer. Analyt. Meth. Geo-
mech., 24共5兲, 509–529.
Hardy, G. H. 共1994兲. Divergent series, Oxford University Press, Oxford,
U.K.
Kramer, S. L. 共1991兲. “Behavior of piles in full-scale field lateral loading
tests.” Report No. WA-RD 215.1, Washington State Dept. of
Transportation, Va.
Matsui, T. 共1993兲. “Case studies on cast-in-place bored piles and some
considerations for design.” Deep foundation on bored and auger piles,
W. F. Van Impe, ed., Balkema, Rotterdam, The Netherlands, 77–101.
Ooi, P. S. K., and Ramsey, T. L. 共2003兲. “Curvature and bending
moments from inclinometer data.” Int. J. Geomech., 3共1兲, 64–74.
Pinto, P. L., Anderson, B., and Townsend, F. C. 共1999兲. “Comparison of
horizontal load transfer curves for laterally loaded piles form strain
gages and slope inclinometer: A case study.” Field instrumentation for
soil and rock, ASTM STP 1358, G. N. Durham and W. A. Marr, eds.,
ASTM, West Conshohocken, Pa., 3–15.
Prakash, S., and Sharma, H. D. 共1990兲. Pile foundations in engineering
practice, Wiley, New York.
Reese, L. C., and Van Impe, W. F. 共2001兲. Single piles and pile groups
under lateral loading, Balkema, Rotterdam, The Netherlands,
121–123.
Reese, L. C., Wang, S. T., Isenhower, W. M., Arrellaga, J. A., and Hen-
drix, J. 共2000兲. Documentation of computer program LPILE Plus ver-
sion 4.0 for Windows, Ensoft, Inc., Austin, Tex.
Wang, J. T. S., and Lin, C. C. 共1999兲. “A method for exact series solutions
in structural mechanics.” J. Appl. Mech., 66共2兲, 380–387.
Wang, S. T., and Reese, L. C. 共1991兲. “Analysis of piles under lateral
load-computer program COM624P for the microcomputers.” Report
No. FHwA-SA-91-002, U.S. Dept. of Transportation, FHwA,
Washington, D.C.